1358244 九、發明說明： 【發明所屬之技術領域】 本發明係有關於將藉焊接處理所形成塡料角之基板， 〜面從既定方向照明一面拍攝，並使用所產生的影像中之 塡料角的正反射光影像自動檢查各塡料角之形狀的基板外 觀檢査裝置（以下，亦有只稱爲「檢查裝置」的情況）。尤 其’本發明係有關於用以將塡料角之檢查所需的檢查基準 資料設定於檢査裝置的技術。 【先前技術】 自動執行基板之外觀檢査的型式之檢查裝置，一般具 備有相機或線感測器等之攝影手段、及被裝入電腦的控制 器。控制器取入藉攝影手段所產生之檢查對象的基板之影 像，並量測塡料角等之被檢查部位的位置或大小，藉由將 所得之量測値和既定之判定基準値進行比較，而判定被檢 查部位之良否。 作爲這種基板外觀檢査裝置，申請人開發有引入稱爲 「彩色強光方式」之光學系的自動外觀檢查裝置。彩色強 光方式係將紅、綠、藍之3種彩色光從各自相異的仰角方 向照射於基板’並拍攝來自被檢查部位（一般爲焊接部位） 之正反射光’而根據紅、綠、藍各色的分布產生表示被檢 査部位之傾斜狀態的影像（參照專利文獻1 )。在具有此彩色 強光方式之光學系的檢査裝置，藉二値化處理而檢測各色 區域’量測其位置或大小等並和判定基準値進行比較，藉 此判定是否正常地形成塡料角。 1358244 [專利文獻1]特公平6 — 1173號公報 藉基板外觀檢查裝置進行自動檢查時，在檢查之前， 需要設定各種檢查基準資料，並登錄於檢查裝置的記憶體。 在本說明書所指之「檢查基準資料」，係表示使用何種 方法並按照何種步驟檢查係檢查對象部位的塡料角。可 說’亦可係表示在進行關於檢查之一連串的處理（影像之產 生、被檢查部位的檢測、量測、判定等）上應遵守之各種基 準的資料。 檢查基準資料包含有檢查對象區域之設定所需的資料 (區域的位置及大小）、爲了檢測被檢查部位而執行之處理 的種類（二値化處理、邊緣抽出處理、投影處理等）、對檢 查對象區域之量測處理的方法、判定量測結果之良否所需 的判定基準値等。又，在藉二値化進行被檢查部位之檢測 的情況，在其處理所使用的二値化臨限値亦包含於檢查基 準資料。 在以往之檢查基準資料的設定處理（所謂的「教導」）， 一般作成拍攝各被檢査部之狀態係良好的模型之基板（以 下稱爲「良品模型」）並顯示，再由相關作業人員一面確認 此顯示之畫面，一面設定各元件所需的檢查基準資料。 又爲了減輕檢查基準資料之設定所花費的勞力，亦有 預先對元件之各種類登錄標準的檢查基準資料（程式庫資 料），再讀出該程式庫資料並設定的情況。例如，在如下之 專利文獻2，記載將從CAD資料所讀出之元件的位置資訊 和程式庫資料組合，而製作檢查基準資料（在專利文獻2記 13582441358244 IX. Description of the Invention: [Technical Field] The present invention relates to a substrate on which a substrate angle formed by a soldering process is applied, and the surface is photographed from a predetermined direction, and the angle of the image in the generated image is used. The substrate reflection inspection device automatically checks the shape of each of the coating angles (hereinafter, also referred to as "inspection device"). In particular, the present invention relates to a technique for setting inspection reference data required for inspection of a corner of a material to an inspection apparatus. [Prior Art] A type inspection device that automatically performs an appearance inspection of a substrate generally has a photographing means such as a camera or a line sensor, and a controller incorporated in the computer. The controller takes in an image of the substrate to be inspected by the photographing means, and measures the position or size of the portion to be inspected such as the angle of the pick, and compares the obtained measured value with the predetermined determination reference And determine whether the part to be inspected is good or not. As such a substrate appearance inspection device, the applicant has developed an automatic visual inspection device which introduces an optical system called "color glare mode". In the color glare mode, three kinds of colored lights of red, green, and blue are irradiated onto the substrate from different elevation directions and the positive reflected light from the inspected portion (generally a welded portion) is taken, and according to red, green, The distribution of the blue colors produces an image indicating the tilted state of the portion to be inspected (see Patent Document 1). In the inspection apparatus of the optical system having the color glare method, the position and size of each color region are detected by binarization processing, and compared with the determination reference ,, thereby determining whether or not the sag angle is normally formed. In the automatic inspection by the substrate visual inspection device, it is necessary to set various inspection reference materials and register them in the memory of the inspection device before the inspection. The "inspection reference data" referred to in this manual indicates which method is used and in which step the inspection angle of the inspection target portion is checked. It can be said that it can also be used to indicate various standards to be observed in a series of processes (image generation, detection, measurement, determination, etc.) of one of the inspections. The inspection reference data includes the data (the position and size of the area) required for the setting of the inspection target area, the type of processing to be performed to detect the part to be inspected (division processing, edge extraction processing, projection processing, etc.), and inspection The method of measuring the target area, the determination criterion required to determine the quality of the measurement result, and the like. Further, in the case where the inspection of the portion to be inspected is carried out by means of the second inspection, the dilatation threshold used in the treatment is also included in the inspection reference data. In the setting process of the inspection standard data (so-called "teaching"), it is generally used to display a substrate (hereinafter referred to as "good model") in which the state of each inspection unit is good, and display it, and then the relevant worker Check the display screen and set the inspection standard data required for each component. In addition, in order to reduce the labor required for setting the inspection reference data, there are cases in which the inspection standard data (database material) of various types of component registration standards are read in advance, and the library data is read and set. For example, in the following Patent Document 2, the positional information of the component read from the CAD data and the library data are combined to prepare inspection reference data (Patent Document 2, 1358244).
載爲「檢查資料」）。 又’在如下之專利文獻3 自動製作塡料角檢查用之檢| 檢查參數作爲程式庫資料並羞 導線寬度之1/2的長度設爲貴 用元件之尺寸資料的計算式， 限値。 [專利文獻2]特開2004 -[專利文獻3]特開平1 1 _ 【發明內容】 〔發明要解決之課題〕 在專利文獻3，雖然對名 作爲程式庫資料並登錄，但是 限値的法則（在專利文獻3稱 作爲程式庫資料並登錄，而右 組裝之元件的形狀資料和程 臨限値。若如此做，因爲對-件種類之各元件登錄共同的杉 定用臨限値之設定法則，或违 況，亦可易於應付。 可是，因爲塡料角形狀， 的乳液焊膏高度而變化之情济 ，記載根據元件的形狀資訊， ί參數（判定用臨限値），將此 :錄。作爲具體例，記載有將 [小塡料角長度等，並根據使 求得塡料角檢查之判定用臨 7 1 7 8 1號公報 311508號公報 元件將具體之判定用臨限値 亦可替代之，將用以導出臨 爲「檢査參數算出方法」者） 教導時，使用實際上基板所 庫資料，算出具體之判定用 種元件種類，可對屬於此元 查基準資料，所以在變更判 加尺寸相異之新的元件之情 因被印刷於焊接前之接線座 ，所以直接使用共同劃一的 檢查基準資料時，具有難確保檢查之精度的可能性。 乳液焊膏高度變動的原因之一，在於在焊膏印刷步驟 1358244 所使用之金屬遮罩的構造（尤其遮罩之厚度）。因爲遮罩係 配合基板之構造而設計，所以即使係同一元件，亦因爲所 組裝之基板，而具有焊料高度變成相異的可能性。 又，因爲塡料角係形成於元件側電極和接線座之間， 在元件之尺寸爲變動大的元件種類之情況，相對於元件高 度之焊料相對高度的差異會影響塡料角形狀。 例如，在相同之基板組裝元件高度相異的2個方形晶 片之情況，裝載各元件之接線座的形狀或大小係相同，對 這些接線座，將乳液焊膏印刷所使用之遮罩的開口部之形 狀或大小被設爲柑同。在使用此遮罩對各接線座乳液印刷 焊膏後，組裝該2個方形晶片，並執行回流焊步驟時，對 於兀件商度比較局之方形晶片，熔化的焊膏鼓起至接近元 件|側之電極的上面之高度爲止，形成陡峭的塡料角，而對 於元件高度比較低之方形晶片，塡料角的傾斜變緩。 本發明係著眼於上述之問題點而開發者，其目的在於 作成即使在塡料角形狀因焊料高度而變成相異的情況，亦 可自動地設定適合實際之塡料角形狀的檢查基準資料。 〔解決課題之方式〕 本發明之方法係在將已形成塡料角之基板一面從既定 方向照明一面拍攝，並使用所產生的影像中之塡料角的反 射光影像來自動檢査該塡料角之形狀的檢查裝置，設定檢 査基準資料的方法。在此方法，對各元件種類，預先製作 用以將對應於彼此相異之塡料角形狀的複數種檢査基準.資 料’和關於各自對應之形狀的塡料角之形成的焊料高度範 1358244 圍賦與關聯並登錄的資料庫。而且，使用基板的設計資料， 特定成爲塡料角檢查之對象的元件，並將所特定之各元件 作爲處理對象，並各自執行以下之步驟A、B、C以及D。 在步驟A，特定和所特定之元件對應的接線座之位置 及大小。在步驟B，對位置及大小已被特定之各接線座， 根據其特定內容來設定塡料角檢查用的檢查對象區域。在 步驟C，對於已被設定檢査對象區域之接線座，取得表示 關於塡料角的形成之焊料高度的資訊，並從和資料庫所登 錄之處理對象的元件對應之元件種類的檢查基準資料之中 讀出和取得的資訊（表示焊料高度之資訊）對應之檢查基 準資料。在步驟D，對於在步驟B所設定之檢查對象區域， 將其設定所需的資訊和在步驟C所讀出之檢查基準資料賦 與對應，並登錄於檢査裝置的記憶體。 在步驟A，雖然亦可作成從基板的設計資料直接讀出 和所特定之元件對應的接線座之位置及大小，但是未限定 如此。例如’亦可從拍攝乳液焊膏之印刷前的裸基板而得 之影像檢測基板上之各接線座，並對其中最接近處理對象 之元件的位置之接線座，量測位置及大小。進行此處理的 情況之元件的位置，可從基板之設計資料求得，亦可受理 由使用者之座標的輸入。 在藉該方法所使用的資料庫，較佳爲至少對塡料角應 執行之量測處理的內容，或對於判定藉此量測處理所求得 之量測値的適當否所需的判定用臨限値，登錄根據熟練人 員之知識或經驗的檢査基準資料。又，較佳爲對元件之各 1358244 種類，登錄和在其元件種類可能產生之多種塡料角形狀對 應的檢查基準資料。 若依據上述之方法，可根據基板的設計資料設定檢查 對象區域，而且可根據此檢查對象區域內之接線座的焊料 高度，設定適合由此焊料所形成之塡料角形狀的檢查基準 資料。 在該方法，作爲「關於塡料角之形成的焊料高度」，例 如使用在元件之組裝前被塗布於接線座的乳液焊膏高度。 在此情況，表示焊料高度之資訊’例如可使用在乳液焊膏 印刷步驟印刷狀態係良好的基板’並利用量測乳液焊膏高 度之方法而取得。 較佳爲，將藉回流焊步驟而熔化時之焊料高度作爲「關 於塡料角之形成的焊料高度」。其原因爲’若係相同的基 板，因爲遮罩之開口部的厚度係大致相同’所以乳液焊膏 高度亦變成大致相同’但是執行回流焊步驟時’因爲接線 座的大小、元件高度、焊料之表面張力等的要因，導致焊 料高度發生變化之緣故。雖然在此情況’要量測焊料高度 係困難，但是藉以下之形態所示的計算’可求得大致之値。 在該方法之一形態’在步驟c’從包含有表示焊料印 刷用之遮罩的開口部之大小的資訊之基板設計資料’讀出 接線座的大小a及遮罩之開口部的大小b’並藉由將預先 所輸入之該遮罩的厚度c和該b之乘積除以a之計算 ((bxc)/a)，而算出關於塡料角之形成的焊料高度。 更佳之形態爲，在步驟C’從基板之設計資料讀出遮 -10- 1358244 胃之開口部的大小b，而且使用相同之設計資料及處理對 象之元件的形狀資料’求得接線座之未裝載元件的部分之 大小al，並藉由將預先所輸入之該遮罩的厚度c和該b之 乘積除以al的計算（（bxc)/a1)’而算出焊料高度。 在上述2種形態的計算，係相當於假設被埋入遮罩之 開口部的乳液焊膏熔化，且均勻地擴散至接線座，求得此 熔化時之焊料高度。若依據這些形態，即使無實際之基板’ 亦可利用計算求得乳液焊膏高度，所以檢查基準資料的設 定變得容易。 此外，任一種形態，亦都考慮在乳液焊膏熔化時助熔 劑蒸發而體積減少的現象，而亦可將對利用上述之計算所 得的値乘以既定之收縮率的値，作爲焊料高度。 在將特定之元件種類的元件作爲處理對象之情況的較 佳形態爲，在步驟C，從已登錄此元件種類所含之各元件 的高度之元件資料庫讀出在步驟A所特定的元件之高度， 並取得相對於所讀出之元件的高度之焊料的相對高度，作 爲表示關於塡料角之形成的焊料高度之資訊。 若依據此形態，對於如方形晶片也之尺寸的變動大之 元:件種類，亦可根據相對於所裝配之元件之焊料的相對高 度’設定適當之檢查基準資料。 此外，在本發明，作成在即使關於塡料角之形成的焊 料高度係大致相同，亦因其他的要因而需要設定相異之檢 査基準資料的情況，亦可應付。 &進行此設定之情況的一形態，對於具有在塡料角的 -11- 1358244 射度應根 反高對於 次 一自屬 二同各將 之之況 ， 角料情 C 料焊之驟 塡在生步 的，發在 件類未 ， 元種及又 之件況。 鄰元情庫 相的的料 自 性射資 來能反於 有可次錄 含之二登 包像生料 中影發資 像光和準 影的將基 光起，查 射引圍檢 反所範的 據各二次反射的有無而已登錄檢查基準資料之元件種類的 元件作爲處理對象，且在和所取得之塡料角的高度對應之 檢查基準資料存在複數個時，使用基板的設計資料並算出 處理對象之元件的接線座和面對該接線座的相鄰元件之接 線座之間的距離。而且，根據算出之距離和既定之臨限値 的大小關係，決定從資料庫要讀出之檢査基準資料。 若依據上述之形態，雖然在實際的塡料角形狀無問 題，但是對因爲二次反射而具有塡料角之反射光影像變成 和一般相異的狀態之可能性的元件種類（例如方形晶片）， 可自動地設定已考慮到二次反射之影響的檢査基準資料。 又，在其他的形態，在資料庫，對於具有塡料角的形 狀根據電極部分之濕潤性而變化的可能性之元件種類，在 焊料之同一高度範圍，登錄和濕潤性良好之情況及不良之 情況之各自對應的檢查基準資料。又，在步驟c，作成將 屬於根據各濕潤性的良好、不良而已登錄檢查基準資料之 元件種類的元件作爲處理對象，且在和所取得之塡料角的 高度對應之檢査基準資料存在複數個時，取得關於此元件 之濕潤性的良好、不良之資訊，並從資料庫讀出和所取得 之資訊對應的檢查基準資料。 若依據上述之形態，對於即使在焊料高度無大的差 -12- 1358244 異，亦因爲塡料角的形狀受到電極部分之濕潤性而變化的 元件種類（例如1C)，亦可切換根據濕潤性之良否而設定的 檢查基準資料。 作爲取得關於濕潤性之良好、不良的資訊之方法，有 受理使用者之輸入的方法。又，預先製作已登錄各元件之 電極材料的電極材料資料庫，及將各種電極材料和濕潤性 之關係賦與對應的比對用表，並對處理對象之元件，在使 用電極材料資料庫特定電極材料後，根據此已特定之電極 材料和比對用表進行比對，藉此亦可取得關於濕潤性的良 好、不良之資訊。 執行上述之檢查基準資料的設定方法之基板外觀檢查 裝置具備有：資料庫，係對各元件種類，將對應於彼此相 異之塡料角形狀的複數種檢查基準資料，和關於各自對應 之形狀的塡料角之形成的焊料高度範圍賦與關聯並登錄； 檢查基準資料設定手段，係使用檢查對象之基板之設計資 料’設定適合檢査對象之基板上的各元件之檢查基準資 料；以及記憶體，係用以保存藉檢查基準資料設定手段所 設定之檢查基準資料。檢查基準資料設定手段具備有以下 之各手段：元件特定手段，係特定成爲塡料角檢查之對象 的元件及其元件種類；接線座特定手段，係對於元件特定 手段所特定之元件’特定和該元件對應之接線座的位置及 大小；區域設定手段，係對已被特定位置及大小之各接線 座’根據其特定內容來設定塡料角檢查用的檢查對象區 域；資料抽出手段’係對於已被設定檢查對象區域之接線 -13- 1358244 座，取得表示關於塡料角的形成之焊料高度的資 該資料庫所登錄之處理對象的元件之檢查基準資 出和所取得的資訊對應之檢查基準資料：以及登 對於藉區域設定手段所設定的檢查對象區域，將 設定所需的資訊和資料抽出手段所讀出之檢查基 與對應，並登錄於記億體。 在上述之基板外觀檢查裝置，雖然已登錄檢 料之資料庫被保存於基板檢查裝置的記憶體內者 但是未限定如此，亦可作成使用CD - ROM等之 記憶媒體，藉由和外部機器的通信，而提供基板 裝置。 〔發明之效果〕 若依據上述之檢查基準資料的設定方法、及 法之基板外觀檢查裝置，在使用資料庫自動地設 準資料的情況，可，根據關於塡料角之形成的焊料 定適合塡料角形狀之檢查基準資料。因而，大幅 檢查等準資料之自動設定處理的性能，並可使用 檢查基準資料執行高精度的檢查 【實施方式】 第1圖係表示應用本發明之基板外觀檢查 造。 此基板外觀檢査裝置（以下，僅稱爲「檢查g 將經由回流焊步驟之基板作爲處理對象，檢查各 接部（塡料角）者，由控制器1、相機2、照明部3 訊，並從 料之中讀 錄手段， 此區域之 準資料賦 查基準資 爲較佳， 可移動式 外觀檢查 應用此方 定檢査基 高度，設 度地提高 所設定之 裝置的構 t置」）係 元件的焊 、基板工 -14- 1358244 資料庫101、102所產生之檢査資料檔案103等 料檔案103，對各元件登錄元件之塡料角檢查 基準資料。此外’在將基板8分成多個區域並拍 在檢查資料檔案103中，亦登錄將相機2之視 影對象位置所需的資料（基板工作台之移動量等 控制部1 0藉由經由XY工作台控制部1 4 構42之移動，而將相機2和基板8進行位置對 藉此攝影所產生之彩色影像係經由影像輸入部 入控制部10，並儲存於其內部的記憶體（RAM | 10對此RAM所儲存之彩色影像的各元件，使 於檢查資料槽案103中之檢查基準資料，依序 角的檢查。 此外，控制部10使用通信用介面17向未 處理裝置傳送對各元件的量測結果或判定結果 使用之影像。 這種檢査裝置，多數情形係作爲教導處理 件之塡料角處於良好的狀態之模型的基板，並 影所產生之影像製作檢查基準資料。可是，在 檢查裝置，係作成使用檢查基準資料庫101、 種元件之形狀的元件形狀資料庫102、基板之 等，自動地設定各種檢査基準資料，而不拍攝模 爲了此設定，在檢查裝置之控制器1，裝入如1 般的檢査基準設定系統1〇〇。 在此檢査基準設定系統1〇〇，除了上述之 。在檢查資 所需的檢查 攝的情況， 野對準各攝 )° 控制移動機 準並拍攝。 1 1而被輸 I)。控制部 用各自登錄 執行對塡料 圖示之資訊 、及檢查所 ，拍攝各元 使用藉此攝 本實施例的 已登錄有各 CAD資料 :型的基板。 窘2圖所示 檢查基準資 -16- 1358244 料庫101、元件形狀資料庫102、以及檢查資料檔案i 03以 外，還包含有CAD資料記億部104、元件、接線座識別部 1〇5、視窗設定部106、焊料高度算出部107、檢查基準資 料檢索部108、臨限値算出部109、檢査基準登錄部H〇、 以及判定基準表1 1 1等。 上述的檢查基準設定系統100之中，CAD資料記憶部 104、檢查基準資料庫101、元件形狀資料庫102、檢查資 料檔案1 03、以及判定基準表1 1 1係各自被設定於控制器1 之記憶體15內。其他的部分係藉安裝於記憶體15之程式 所設定於控制部1 0的功能。 此檢查基準設定系統100所讀入的CAD資料，用於特 定基板所裝載之各元件及其元件種類的處理、或用於特定 對應於各元件之接線座的位置及大小之處理。具體而言， 關於各元.件’設定用以表示組裝位置、元件種類名稱（方形 晶片、1C、電容器等之一般名稱）、型號等之資訊，且設定 用以表示接點座之位置及尺寸之資訊（以下，稱爲「接線座 資訊」）。又，亦設定表示在焊膏印刷步驟所使用之金屬遮 罩的開口瓿之位置及大小的資訊（以下稱爲「遮罩資訊」）。 此遮罩資訊雖然本來係在乳液焊膏印刷步.驟所使用者，但 是在本實施例，係由焊料高度算出部10 7使用。 上述的CAD資料，係從被設定於CD — ROM驅動器16 的CD-ROM所讀出，或從未圖示的外部裝置所傳送，再藉 通信用介面17取入，並被儲存於CAD資料記憶部1〇4。， 檢查基準資料庫101及元件形狀資料庫102係預先被 -17- 1358244 安裝於檢查裝置。但’任一個資料庫101、102都可藉CD - ROM或和外部之通信而適當地更新。 在檢查基準資料庫101，其細節將後述，係對各元件 種類’登錄和在此元件種類可產生之多種塡料角形狀對應 的檢查基準資料。資料庫中之元件種類，係以和CAD資料 的元件種類匹配的方式被分類，對各元件種類，賦與和在 CAD資料所使用的相同之元件種類名稱。 在對應於一個塡料角形狀的檢查基準資料中，包含有 用以指定從塡料角之影像所檢測的顏色之資料、用以指定 在量測或判定所執行的程式之資料（例如，程式之檔案名稱 或表示儲存位置的位址等）、用以求得判定用臨限値的法則 (以下，稱爲「臨限値導出法則」）等》 元件形狀資料庫102係對各個元件儲存型號、元件種 類名稱、廠商名稱、型式等用以特定其元件之資訊，和元 件的形狀資料。形狀資料係藉具體之數値表示元件的尺寸 或形狀者，對於任一元件，都儲存有元件本體之縱橫的寬 度或元件之高度。此外，亦有對元件種類儲存特有之資料 的情況。例如，若係屬於「1C」之元件，則儲存導線間的 間距、各個導線之寬度、相對向的導線之端邊間的距離、 導線之高度等。 元件、接線座識別部1〇5係根據CAD資料中的元件種 類名稱或型號（以下，將這些稱爲「元件資訊」），逐一分 割被組裝於基板之元件並識別。此外，對各元件，特定對 應於其元件之接線座，並取得各接線座的接線座資訊。視 -18- 1358244 窗設定部106係從元件、接線座識別部1〇5取入對應 •元件之接線座資訊，並根據此資訊，設定包含有接線 大小的檢查對象區域（以下，稱爲「接線座視窗」）。 從元件、接線座識別部105將元件資訊及接線座 供給檢査基準資料檢索部108。檢查基準資料檢索部 從這些資訊識別教導對象之元件的元件種類或接線庳 小時，對各接線座，從檢査基準資料庫101所登錄的 / 之檢查基準資料之中，特定適合形成於其接線座的塡 形狀，並讀出其資料。在用以特定此檢查基準資料之處 使用判定基準表111。又，除了 一部分之元件以外， 料高度算出部107執行算出接線座上之高度的處理。 檢查基準資料檢索部108從檢査基準資料庫101 特定之檢查基準資料時，將此資料中的臨限値導出法 給臨限値算出部109»臨限値算出部109藉將元件之 資料應用於此臨限値導出法則的計算處理，而算出對 對象之元件所特定的塡料角檢查用之判定用臨限値。 檢查基準登錄部110分別取入包含有該判定用臨 之各種檢查基準資料，而且從視窗設定部106接受接 視窗之設定所需的資訊（係表示視窗之位置及大小 訊。在以下，稱爲「接線座視窗之設定資訊J )之供給 這些資訊賦與對應，並儲存於檢查資料檔案1〇3。 此外，在檢査資料檔寒103，作爲各元件之共同 查基準資料，亦儲存檢測紅、綠、藍之各顏色所需的 化臨限値。此二値化臨限値亦被登錄於檢查基準資 於各 座之 資訊 108 之大 複數 料角 .理， 使焊 讀出 則供 形狀 教導 限値 線座 的資 ，將 的檢 二値 料庫 -19- 1358244 10卜利用未圖示之處理部讀出並儲存於檢查資料檔案10 3。 又，在上述之例子，雖然CAD資料採用包含有表示元 件種類的資訊，但是在使用未含有元件種類資訊之CAD資 料的情況，則係另外由使用者輸入元件種類資訊，並將此 輸入資訊供給元件、接線座識別部105。 以下，以係主要之元件種類的「方形晶片」及「1C」 爲例，說明檢查基準資料之具體的內容或設定所需之處理。 第3圖係對於該2種元件種類，表示檢查基準資料庫 1〇1所登錄之資料的構造例。在此例，對於各元件種類， 將在其元件種類所產生之塡料角形狀分類成複數種型式 (以下稱爲「塡料角型式」），並對各塡料角型式各個設定 個別的檢查基準資料。 在此實施例’爲了使用者之方便，雖然以「方形晶片 A」、「IC_A」等包含有元件種類名稱的名稱表示各塡料角 型式’但是塡料角型式不是用以將元件分類者，而係用以 將塡料角形狀分類者。即，在屬於同一元件種類之任一元 件’亦具有形成和其元件種類對應之全部的塡料角型式之 塡料角的可能性。又，如後述所示，即使係相同之元件， 亦有因接線座而塡料角型式相異的情況。 在檢查基準資料，雖然包含有上述之多種資料，但是 在第3圖，限定爲臨限値導出法則並圖示。又，在近年來 之彩色強光方式的檢査裝置，雖然一定檢測藍色區域，但 是因爲因應於需要檢測綠、紅色區域，所以在本例，亦權 宜上’當作僅量測藍色區域，並表示對藍色區域之量測値 -20 - 1358244 的臨限値導出法則。 第3圖所示的資料之 資料係塡料角型式之名稱 狀J或「影像」係表示各 錄於檢查基準資料庫101 照，亦可保存於檢查基準 「塡料角形狀」係填米 係藉彩色強光方式之光學 圖，對應於塡料角之部分 色成藍或紅（在第3圖， 域，利用網點之圖案表示 以省略著色表示）。 作爲「臨限値導出法丨 料角之寬度及長度（以下ί 度」）的量測値求臨限値之 外，因爲設定對元件寬度 則，所以根據元件而設定 第4圖表示在方形晶 槪念。 方形晶片之情況的^ : 向正交的方向（圖中之縱） 件寬度方向之藍色區域的 在和元件寬度方向正交之 於1C的情況，雖未圖示， 中’檢查基準資料庫101所需的 和臨限値導出法則。「塡料角形 型式之具體的內容者，未必要登 。例如，爲了使用者可適當地參 資料庫101以外之檔案。 午角之截面形狀的模式圖。「影像J 系所得的影像中之塡料角的模式 ，係配合實際的顏色分布，被著 分別利用斜線之圖案表示藍色區 紅色區域。因爲綠色區域小，所 則」，在此實施例，設定用以對塡 爯爲「塡料角寬度」、「塡料角長 法則。又，除了一部分的型式以 或導線寬度乘以既定之係數的法 相異之臨限値。 片之塡料角寬度及塡料角長度之 元件寬度」，係在和電極之排列方 ί向）之寬度。塡料角寬度係在元 長度之最大値a，塡料角長度係 藍色區域的長度之最大値b。關 但是一樣地，將在沿著導線之寬 -21 - 1358244 度的方向之藍色區域的長度之最大値作爲塡料角寬度，而 將在沿著導線之長度方向的藍色區域之長度的最大値作爲 塡料角長度。 在此，說明各元件種類之塡料角形狀和臨限値導出法 則的關係。 首先，關於「方形晶片」，雖然設定4種塡料角型式， 但是方形晶片A和C之塡料角形狀係大致相同。在這些型 式，以元件之比較高的位置（在圖示例爲上端緣）爲起點， 形成陡峭之塡料角。 方形晶片B在塡料角長度，雖然和方形晶片A或C之 大致相同，但是因爲塡料角的起點係位於元件之比較低的 位置，所以塡料角之傾斜比方形晶片A、C更緩。因而， 在塡料角之影像，在前端側出現紅色區域。 方形晶片A、C係根據二次反射之有無而區分。二次 反射意指將來自位於相對向關係之相鄰的元件之塡料角的 反射光照射於塡料角，而發生對此照射光之反射的狀態。 已知在塡料角係陡峭，且和相鄰之塡料角的距離短之情 況，易發生二次反射。在此實施例，將未發生二次反射之 型式設爲方形晶片A，將發生二次反射的型式設爲方形晶 片C。雖然對應於方形晶片A之塡料角的影像，大致整體 變成藍色，但是在對應於方形晶片C之塡料角的影像，在 藍色區域之內部出現二次反射所引起的紅色區域。 相對於上述之3種型式，方形晶片D因爲接線座極小， 所以對應於形成短且陡峭的塡料角。此型式之塡料角的影 -22- 1358244 像’係和方形晶片A —樣，大致整體變成藍色。但，塡料 角長度，係遠比方形晶片A的小。 在各型式的臨限値導出法則，反映上述之影像的差 異。具體而言，在方形晶片A，將相當於元件寬度之0.5 倍的數値設定爲塡料角寬度的臨限値，並將相當於元件寬 度之0.6倍的數値設定爲塡料角長度的臨限値。相對於此， 在方形晶片B，雖然將塡料角寬度的臨限値設定爲和方形 晶片A相同，但是關於塡料角長度之臨限値，考慮紅色區 域的存在，設定爲比方形晶片A更小之値。又，關於方形 晶片C，考慮二次反射之發生，塡料角寬度、長度都設定 爲比方形晶片A小的臨限値。 關於塡料角陡峭且短的方形晶片D，在塡料角寬度之 臨限値，雖然設定爲和方形晶片A相同，但是將塡料角長 度之臨限値固定爲定値（20 μιη)。關於這種塡料角，因爲在 經驗上已知變成大致相同的形狀，和成爲塡料角之前的乳 液焊膏之高度無關。 其次，在「1C」，3型式之中的1(：_八相當於陡峭之塡 料角。在此型式，和方形晶片Α —樣，產生塡料角之大致 整體變成藍的影像。 相對於此，對應於IC_B的塡料角，以電極之比較低的 位置爲起點，緩慢地傾斜。在此型式之塡料角的影像，紅 色區域變得顯著。尤其，在對應於塡料角之前端緣的部分， 寬度方向的大部分變成紅色。 對應於IC_C的塡料角，變形成中央部平坦而傾斜部分 -23 - 1358244 少的形狀。這種塡料角之變形係在導線表面形成鈀的電鍍 - 層等，而導線之濕潤性（對導線之焊料的安裝狀態）變差的 ~~. .情況發生。在此型式之塡料角的影像，沿著長度方向出現 紅色區域。 同樣地在「1C」用的臨限値導出法則，亦作成因應於 如上述所示之影像的差異，而各型式之判定用臨限値相 異。具體而言，關於IC_B或IC_C，作成塡料角寬度之判 定用臨限値比IC_A的小。又，關於塡料角長度之臨限値， 雖然在IC_A或IC_B，設定定値（50μιη、30μιη)，但是關於 IC_C，臨限値因應於接線座之長度而變化。 在判定基準表111，爲了判定教導對象之元件的塡料 角屬於上述之哪一種型式，而儲存如第5圖所示的內容之 資料。其中，「影像」欄之模式圖係供參考之表示，亦可不 儲存於實際的表。 在此判定基準表111，將檢查基準資料庫101所設定 的塡料角型式設定爲對關於塡料角之形成的焊料之高度 (以下只稱爲「焊料高度」）賦與對應的資料。方形晶片A、 B、C的焊料高度各自以焊料高度對元件高度之百分比的數 値範圍表示’ IC_A、B、C的焊料高度係以具體之數値的範 圍表示。另一方面，關於亦可不考慮焊料高度之方形晶片 D，在焊料高度儲存零（nun)値。 此外，在方形晶片之4型式的資料，包含有對應之接 線座的長度（接線座長度）。又，在方形晶片八及c之資料， 包含有用以判別—次反射之有無的參數（形成塡料角之接 -24 - 1358244 線座和與其相對向的相鄰之元件的接線座之距離）。又，IC 中，亦在IC_A及IC_C的資料，包含有表示導線之濕潤性 的良好、不良之資料。 第3圖所示之檢查基準資料庫101和第5圖所示的判 定基準表111係經由塡料角型式之名稱而連結。藉此連 結’除了方形晶片D以外之各型式的檢查基準資料，對關 於各自對應之形狀的塡料角之形成的焊料高度範圍被賦與 對應。因而，若知道實際的接線座所形成之焊料高度，就 可特定和其高度對應的檢查基準資料。 又，在方形晶片A和C，以及1C之A和C，雖然被賦 與對應同一之高度範圍，但是依據第5圖，方形晶片A和 方形晶片C可根據和相對向接線座之距離而區分，IC _ A和 IC_C可根據濕潤性之良好、不良而區分。又，焊料高度未 被賦與對應之方形晶片D，亦根據接線座長度而可和其他 的3種型式區分。 第6圖係表示判定方形晶片之塡料角型式的情況之處 理的步驟，第7圖表示判定1C之塡料角型式的情況之處理 的步驟。以下，一面參照第5圖，一面按照各圖之流程， 說明判定處理的細節。此外’在各圖及以下之說明，將各 處理的步驟簡稱爲ST。 在第6圖之方形晶片用的判定處理，首先，對於著眼 中之接線座，從CAD資料中的接線座資訊取得接線座長 度，並檢査此是否是100^m以上（ST101、1〇2)。 依據第5圖’方形晶片A、B、C都對應於ι〇〇μηι以上 -25 - 1358244 之接線座長度，但是方形晶片D對應之接線座長度係小於 ΙΟΟμιη。因而，在ST102之判定爲「NO」的、青況，前進到 ST108 ’並判定塡料角型式係方形晶片d。 另一方面’在接線座長度係ΙΟΟμιη以上的情況，從 ST102進到ST103，並使用焊料高度算出部1〇7之功能， 利用以下之計算處理而算出焊料高度。 首先，使用CAD資料中之接線座資訊或元件資訊求得 接線座的面積或元件之中心位置 '，而且從元件形狀資料庫 102讀出元件本體或電極的尺寸等之資訊，並使用這些資 訊，求得接線座上之焊料印刷對象區域（係未和電極重疊之 部分）的面積SA。此外，從CAD資料抽出遮罩之開口部的 面積SB，而且另外從使用者受理遮罩之厚度δ的輸入，並 使用這些，執行如下的計算式（1)。 焊料高度=(SBxS)/SA ......... (1) 在上述的式（1)’係假設被埋入遮罩之開口部的乳液焊 膏熔化，而均勻地擴散至和接線座之電極未重疊的部分整 體，而求得熔化之焊料的高度。其中，乳液焊膏係將粒狀 的焊料裝入助熔劑者，熔化後，因爲助熔劑的蒸發而體積 減少，所以亦可將因應於此減少量之焊料的收縮率乘以式 (1)之計算結果者，作爲焊料高度。 在接著之ST1 04，從元件形狀資料庫102讀出教導對 象之元件的高度，並檢査在ST103所算出之焊料高度是否 爲元件高度的1/3以上。依據第5圖，對應於方形晶片a、 C之焊料高度係元件高度的1 /3以上，但是對應於方形晶 -26 - 1358244 元B之焊料高度係小於元件高度的1/3。因而，若STl 04 之判定變成「NO」，進到ST 109，並判定塡料角型式係方形 晶片B。 在焊料高度係元件高度之 W3以上的情況，進到 ST105。在ST105，使用CAD資料中之接線座資訊，算出 和位於相對向關係之相鄰的元件之接線座的距離。又在 ST106，將所算出之距離和既定値LM比較。 依據第5圖，雖然在未發生二次反射的方形晶片A， 係和相對向之接線座的距離爲LM以上，但是在發生二次 反射的方形晶片C，此距離爲小於LM。因而，若在ST103 所算出之距離係LM以上，在ST106變成「YES」，並進到 ST 1 07，判定塡料角型式係方形晶片A。另一方面，若所算 出之距離係小於LM，進到S T 1 1 0，並判定塡料角型式係方 形晶片c。 接著，在1C用之判定處理，算出焊料高度後，檢查此 高度是否是0.05mm以上（ST201、202)。依據第5圖，雖然 IC_A及IC_C對應於〇.〇5mm以上之焊料高度，但是對應 於1(：_8之焊料高度係小於〇.〇51!1111。因而，在3丁202爲「：^0」 的情況係進到ST205，並判定塡料角型式係IC_B。 若焊料高度係〇 . 〇 5 m m以上，進到S T 2 0 3，並檢查濕潤 性之良好、不良。此檢查所需之資料係由使用者預先輸入。 例如，顯示詢問對裝載於基板之1C是否施加記的電鑛處理 之判定的畫面，並將對此詢問所輸入之回答使用於ST203 的判定。在此’在判定爲濕潤性良好的情況，進到ST204 , -27- 1358244 並判定塡料角型式係IC_A。而在判定濕潤性不良好的情 況，進到ST206，並判定塡料角型式係IC_C。 接著，根據第8圖說明教導處理之全步驟。 首先，在最初之ST1，讀入教導對象之基板的CAD資 料。接著在ST2，受理對於此基板之關於在焊料印刷步驟 所使用的遮罩之厚度、或1C之導線的濕潤性之良否的資訊 之輸入》 在ST3,製作檢查資料檔案103。其中，此階段的檢查 資料檔案103係僅設定有檔案名稱之空的檔案。在接著之 ST4，將用以計數元件數的計數器n設爲起始値1，以下， 執行ST5〜1 4的迴路。 在ST5,對於根據計數器η所特定之元件（以下稱爲「著 眼元件」），使用CAD資料，特定元件種類名稱或型號。 又在ST6，根據在ST5所特定之型號，從元件形狀資料庫 1 〇 2讀出著眼元件之元件形狀資料。所讀出之元件形狀資 料，係被保存於控制部10內的工作記憶體（RAM)。 接著，在ST7，再使用CAD資料，特定和著眼元件對 應之接線座的位置或大小。雖然在第8圖未明記，但是一 般在ST 7，因爲特定有多個接線座，所以對所特定之每個 接線座，執行以下的ST8〜12。 在ST8，對所特定之各接線座，設定包含有其接線座 之大小的接線座視窗。在ST9，對於各接線座，執行因應 於著眼元件之元件種類的塡料角型式之判定處理（若係方 形晶片爲第6圖的處理，若係1C爲第7圖的處理）。在 -28- 1358244 ST10，對於各接線座，讀出和在ST9所判定 對應的檢查基準資料。 在ST11，對於各接線座，從元件形狀資 度或導線寬度的數値，並將這些値套入應用 之檢查基準資料中的臨限値導出法則，並算 臨限値。在ST12，將包含有此判定臨限値 料’對接線座視窗的設定資訊賦與對應，並 料槍案1〇3。在此階段，對所特定之各接線 準資料和接線座視窗的設定資訊組合，製作 的集合對應於一個元件之最終形態的檢查基 錄於檢查資料檔案103。 然後，在ST 1 3將計數器n增加1。以下 總元件數之値爲止，重複ST5〜12的處理》 在上述的順序，因爲ST8〜11之各步驟 各接線座執行，所以即使係同一元件，亦可 所應用之檢查基準資料相異的情況。例如，屢 在一方之塡料角發生二次反射，而在另一方 生二次反射的情況，對前者設定方形晶片C 料，而對後者設定方形晶片Α的檢查基準資 若依據以上所說明之檢查基準資料的設 件之焊料的相對高度根據元件尺寸之大小 狀.、大小之差異等而變化，對於設想因此變 種塡料角形狀的元件種類，對教導對象之 座’可自動地.設定因應於其接線座所形成之 之塡料角型式 料讀出元件寬 在ST9所讀出 出具體之判定 之檢査基準資 登錄於檢查資 座，將檢查基 使各組合資料 準資料，並登 ，至此η超過 係對所特定之 應付因接線座 i於方形晶片， 之塡料角未發 之檢查基準資 料。 定方法，對元 或接線座的形 化而相異之多 元件的各接線 塡料角形狀的 -29 - 1358244 / 檢查基準資料。又，即使係同一元件，亦若塡料角形狀因 被裝載之基板而異，可設定因應於其形狀的檢查基準資料。 此外，在上述的方法’雖然焊料高度無差異，但是在 因二次反射而在塡料角之影像發生差異，或因電極的濕潤 性而在塡料角形狀發生差異的情況，亦可設定因應於各自 之塡料角的狀態之檢查基準資料。因而，可使用所設定之 檢查基準資料執行高精度的塡料角檢查。 此外，在上述的實施例，雖然作成利用計算算出焊料 高度，但是在乳液焊膏印刷步驟，進行對各接線座所塗布 之乳液焊膏的檢查之情況，亦可對於全部之接線座的焊料 塗布狀態係良好之基板，輸入在焊料印刷檢查所量測的焊 料高度’並將此輸入値或對輸入値乘以既定之係數而得之 數値’用作關於塡料角的形成之焊料高度（係數係可根據上 述之助熔劑的蒸發所伴隨之收縮率等而決定）。 又’在上述的實施例’雖然作成將用以檢測塡料角部 分之顏色的二値化臨限値作爲共同之檢査基準資料並設 定’但是根據元件，亦有在塡料角部分發生相鄰元件之陰 影’而用共同之二値化臨限値無法應付的情況。在可能發 生這種狀況的情況’最好特定從各元件間之距離或位置關 係而發生陰影的可能性高的塡料角，並在此塡料角設定其 他的二値化臨限値。 【圖式簡單說明】 第1圖係表示本發明之一實施例的基板外觀檢查裝置 之構造之圖。 -30- 1358244 第2圖係表示檢查基準設定系統之功能方塊圖。 第3圖係表示檢查基準資料庫之構造例的說明圖。 第4圖係說明元件寬度、塡料角寬度、以及塡料角長 度之槪念的圖。 第5圖係表示判定基準表之構造例的說明圖。 第6圖係表示方形晶片用之塡料角型式的判定處理之 流程之流程圖。 第7圖係表示1C用之塡料角型式的判定處理之流程的 流程圖。 第8圖係表示對一片基板之教導處理的流程之流程 tm 圖。 【元件 符號 說 明 ] 1 控 制 器 2 相 Μ 機 3 照 明 部 10 控 制 部 15 記 憶 體 100 檢 查 基 準 資 料 庫 102 元 件 形 狀 資 料 庫 103 檢 查 資 料 檔 案 104 C AD 資 料 記 憶 部 105 元 件 % 接 線 座 識別部 106 視 窗 設 定 部 1 07 焊 料 高 度 算 出 部 1358244 108 檢查基準資料檢索部 110 檢查基準登錄部 111 判定基準表It is contained in "Check Data". Further, in the following Patent Document 3, the inspection for the inspection of the angle of the material is automatically performed. The inspection parameter is used as the library data and the length of 1/2 of the wire width is set as the calculation formula of the size data of the expensive component. [Patent Document 2] JP-A-2004- [Patent Document 3] Japanese Unexamined Patent Application Publication No. JP-A----- (In Patent Document 3, it is called as a library material and is registered, and the shape data of the right-assembled component and the process limit are limited. If this is done, the setting of the common cedar setting for each component of the type of the component is set. The rule, or the violation of the condition, can be easily dealt with. However, because of the shape of the angle of the material, the quality of the emulsion solder paste changes, according to the shape information of the component, ί parameter (determination threshold), this: As a specific example, it is described that the length of the small nipple angle and the like are determined based on the judgment of the smear angle of the slab. In addition, when the teaching is to be used as the "inspection parameter calculation method", the actual substrate type data is used to calculate the specific type of the component to be used for the determination. When the determination of different size plus a new situation by being printed on the elements of the wire holder prior to the welding, it is common to use direct uniform inspection standard information, a possibility is difficult to secure the accuracy of inspection. One of the reasons for the high variation of the emulsion solder paste is the construction of the metal mask used in the solder paste printing step 1358244 (especially the thickness of the mask). Since the mask is designed to match the structure of the substrate, even if it is the same component, there is a possibility that the solder height becomes different due to the assembled substrate. Further, since the angle of the coating is formed between the element-side electrode and the terminal block, the difference in the relative height of the solder with respect to the height of the element affects the shape of the angle of the material in the case where the size of the element is a variable type of element. For example, in the case of two square wafers in which the same substrate assembly elements are different in height, the shape or size of the terminal blocks on which the components are mounted are the same, and the opening portions of the masks used for printing the emulsion solder paste are applied to the terminal blocks. The shape or size is set to citrus. After the solder paste is printed on each of the terminal blocks by using the mask, the two square wafers are assembled, and when the reflow soldering step is performed, the molten solder paste is swelled to the adjacent component for the square wafer of the component. A steep throw angle is formed up to the height of the upper surface of the electrode, and the tilt of the throw angle is slowed for a square wafer having a relatively low element height. The present invention has been made in view of the above problems, and an object of the present invention is to automatically set an inspection reference material suitable for an actual angle of the corner shape even when the shape of the corner of the material becomes different depending on the height of the solder. [Means for Solving the Problem] The method of the present invention automatically detects a corner of a substrate on which a substrate angle has been formed while illuminating from a predetermined direction, and uses a reflected light image of a rake angle in the generated image. The inspection device of the shape sets the method of checking the reference data. In this method, for each component type, a plurality of inspection standards are prepared in advance to correspond to the shape of the dip angles which are different from each other. The material 'and the height of the solder height 1358244 regarding the formation of the respective angles of the corresponding shapes are assigned to the associated database. Further, using the design data of the substrate, the components to be subjected to the inspection of the angle of the material are specified, and each of the specified components is treated, and the following steps A, B, C, and D are performed. In step A, the position and size of the terminal block corresponding to the particular component is specified. In step B, for each of the terminal blocks whose position and size have been specified, the inspection target area for the inspection of the angle of the material is set in accordance with the specific content. In the step C, the information indicating the height of the solder on the formation angle is obtained for the terminal block to which the inspection target area is set, and the inspection reference data of the component type corresponding to the component of the processing target registered in the database is acquired. The inspection reference data corresponding to the information read and obtained (indicating information on the height of the solder). In step D, the information to be set in the inspection target area set in step B is associated with the inspection reference data read in step C, and is registered in the memory of the inspection apparatus. In the step A, it is also possible to directly read out the position and size of the terminal block corresponding to the specific component from the design data of the substrate, but the limitation is not limited thereto. For example, it is also possible to measure the position and size of each terminal block on the image detecting substrate from the bare substrate before the printing of the emulsion solder paste, and to the terminal block closest to the position of the component to be processed. The position of the component in the case of performing this processing can be obtained from the design information of the substrate, and can also be input by the coordinates of the user. In the database used by the method, it is preferable to determine at least the content of the measurement processing to be performed on the corner of the material, or the determination for determining the appropriateness of the measurement obtained by the measurement processing. After the deadline, log in to the benchmark data based on the knowledge or experience of the skilled person. Further, it is preferable to inspect the reference data corresponding to the various types of material angles which may be generated for the type of each component of the component 1358244. According to the above method, the inspection target area can be set according to the design data of the substrate, and the inspection reference data suitable for the shape of the angle of the pellet formed by the solder can be set based on the solder height of the terminal block in the inspection target area. In this method, as "the height of the solder for the formation of the angle of the feed", for example, the height of the emulsion solder paste applied to the terminal block before assembly of the element is used. In this case, the information indicating the height of the solder can be obtained, for example, by using a method in which the substrate is in a good printing state in the emulsion solder paste printing step and measuring the height of the emulsion solder paste. Preferably, the height of the solder when it is melted by the reflow step is referred to as "the height of the solder formed by the angle of the coating". The reason is that 'if the same substrate is used, since the thickness of the opening of the mask is substantially the same', the height of the emulsion solder paste becomes substantially the same 'but when the reflow step is performed' because of the size of the terminal block, the height of the element, and the solder The cause of surface tension and the like causes a change in solder height. In this case, it is difficult to measure the solder height, but the calculation shown in the following form can be approximated. In one form of the method, in the step c', the size a of the terminal block and the size of the opening portion of the mask are read from the substrate design data "including the information indicating the size of the opening of the mask for solder printing". The height of the solder with respect to the formation of the throw angle is calculated by dividing the product of the thickness c of the mask and the b input in advance by the calculation of a ((bxc)/a). In a preferred embodiment, in step C', the size b of the opening of the stomach is read from the design data of the substrate, and the same design data and the shape data of the component of the processing object are used to obtain the terminal block. The size of the portion of the loading element is a, and the solder height is calculated by dividing the product of the thickness c of the mask and the b, which is input in advance, by the calculation of al ((bxc)/a1)'. The calculation of the above two forms corresponds to the assumption that the emulsion solder paste embedded in the opening of the mask is melted and uniformly diffused to the terminal block, and the height of the solder at the time of melting is obtained. According to these forms, the height of the emulsion solder paste can be calculated by calculation without using the actual substrate, so the setting of the inspection reference data becomes easy. Further, in any of the forms, the flux is evaporated and the volume is reduced when the emulsion solder paste is melted, and the enthalpy obtained by multiplying the enthalpy obtained by the above calculation by a predetermined shrinkage ratio may be used as the solder height. In a preferred embodiment of the component of the specific component type, in step C, the component specified in step A is read from the component database of the height of each component included in the component type. The height, and the relative height of the solder relative to the height of the read element, is used as information indicating the height of the solder with respect to the formation of the angle of the feed. According to this aspect, it is also possible to set an appropriate inspection reference data based on the type of the component such as the square wafer, or the relative height of the solder relative to the mounted component. Further, in the present invention, it is possible to cope with the case where the height of the soldering material is substantially the same even if the height of the solder material is substantially the same, and it is necessary to set the different reference data. & A form of the case where this setting is made, for the -11- 1358244 radiance at the angle of the feed, the radiance of the radiant C is the same as that of the next one. In the case of the birth, it is not in the class, the meta species and the other. The material of the neighboring sentimental phase can be countered by the fact that there are two sub-packages, such as the raw material, the image of the raw material, the light and the shadow, and the investigation. According to the presence or absence of each secondary reflection, the component of the component type of the inspection reference data is registered as the processing target, and when there are a plurality of inspection reference data corresponding to the height of the obtained coating angle, the design data of the substrate is used and calculated. The distance between the terminal block of the component of the processing object and the terminal block of the adjacent component facing the terminal block. Further, based on the calculated distance and the predetermined threshold relationship, the inspection reference data to be read from the database is determined. According to the above-described form, although the actual shape of the angle of the material is not problematic, the type of the element (for example, a square wafer) in which the reflected light image having the angle of the dip is changed to a generally different state due to the secondary reflection. , The inspection reference data that has taken into account the influence of the secondary reflection can be automatically set. Further, in another embodiment, in the data base, the type of the component having the possibility that the shape of the dip angle changes depending on the wettability of the electrode portion is good in registration and wettability in the same height range of the solder, and is inferior. The inspection reference data corresponding to each case. Further, in the step c, an element belonging to the type of the component to which the inspection reference data has been registered according to each of the wettability is formed, and a plurality of inspection reference materials corresponding to the height of the obtained coating angle are present. At the time, good and bad information about the wettability of the component is obtained, and the inspection reference data corresponding to the obtained information is read from the database. According to the above-described form, it is possible to switch the wettability depending on the type of the component (for example, 1C) in which the shape of the twist angle is changed by the wettability of the electrode portion even if the difference in solder height is not large - 12 - 1358244. The inspection reference data set by the good or not. As a method of obtaining good and bad information about wettability, there is a method of accepting input from a user. Further, an electrode material library in which the electrode materials of the respective elements are registered is prepared in advance, and the relationship between the various electrode materials and the wettability is assigned to the corresponding comparison table, and the components to be processed are specified using the electrode material database. After the electrode material is compared, the specific electrode material and the comparison table are compared, whereby good and bad information about the wettability can be obtained. The substrate visual inspection device that performs the above-described method for setting the inspection reference data includes a database, and a plurality of types of inspection reference materials corresponding to the different shape angles of the respective element types, and the respective shapes. Checking the reference data setting means, using the design data of the substrate to be inspected, 'setting the inspection reference data for each component on the substrate to be inspected; and the memory It is used to store the inspection reference data set by the inspection reference data setting means. The inspection reference data setting means includes the following means: the component-specific means specifies the component to be the object of the inspection of the material angle and the component type thereof; the terminal-specific means is specific to the component specified by the component-specific means. The position and size of the terminal block corresponding to the component; the area setting means is to set the inspection target area for the inspection of the angle of the material according to the specific content of each terminal block of the specific position and size; the data extraction means is The wiring of the inspection target area is set to a level of -13,358,244, and the inspection reference standard corresponding to the information to be processed of the component to be registered in the database is obtained. The data and the inspection target area set by the borrowing area setting means are set, and the required information and the inspection base read by the data extraction means are set and corresponding, and registered in the record. In the above-described substrate appearance inspection device, although the database in which the document has been registered is stored in the memory of the substrate inspection device, the present invention is not limited thereto, and a memory medium such as a CD-ROM can be used to communicate with an external device. , and a substrate device is provided. [Effects of the Invention] According to the above-described method for setting the inspection reference data and the substrate appearance inspection device of the method, when the data is automatically set using the database, it is possible to apply the solder according to the formation of the angle of the crucible. Inspection reference data for the shape of the corner. Therefore, the performance of the automatic setting processing of the information can be greatly checked, and the high-precision inspection can be performed using the inspection reference data. [Embodiment] Fig. 1 shows the appearance inspection of the substrate to which the present invention is applied. This substrate appearance inspection device (hereinafter, simply referred to as "inspection g, the substrate to be processed by the reflow process is used as a processing target, and the respective inspection portions (feed angles) are inspected, and the controller 1, the camera 2, and the illumination unit 3 are used. From the material reading method, the quasi-information reference standard in this area is better, the movable visual inspection application uses this to check the base height, and the degree of the device is set to improve the structure of the device. Welding, substrate worker-14-1358244 Data files 101 and 102 generated in the inspection data file 103 and other material files 103, the reference data for the inspection angle of each component registration component. In addition, the board 8 is divided into a plurality of areas and captured in the inspection data file 103, and the information necessary for the position of the camera 2 of the camera 2 is also registered (the control unit 10 such as the movement amount of the substrate table is operated by XY). The movement of the table control unit 14 and the substrate 8 is performed, and the color image generated by the camera 2 and the substrate 8 is input into the control unit 10 via the image input unit, and is stored in the internal memory (RAM | 10). The components of the color image stored in the RAM are inspected by the inspection angle data in the inspection data slot 103. Further, the control unit 10 transmits the components to the unprocessed device using the communication interface 17. The image used for the measurement result or the determination result. In most cases, the inspection device serves as a substrate for teaching the model that the processing angle of the processing member is in a good state, and the image is generated to produce inspection reference data. However, in the inspection The apparatus is configured to automatically set various inspection reference materials without using the inspection reference database 101, the component shape database 102 of the shape of the components, the substrate, and the like. For this setting, the controller 1 of the inspection device is loaded with the inspection reference setting system 1 as shown in Fig. 1. Here, the inspection setting system 1 is checked, except for the above. Situation, the field is aligned with each camera) ° Control the mobile machine and shoot. 1 1 is lost I). The control unit performs the information on the drawing and the inspection site by the respective registrations, and photographs each unit. The substrate on which each CAD data is registered is used. In addition to the inspection reference material-16-1358244, the library 101, the component shape database 102, and the inspection data file i03, the CAD data is recorded in the figure 104, the component, and the terminal identification unit 1〇5. The window setting unit 106, the solder height calculation unit 107, the inspection reference data search unit 108, the threshold calculation unit 109, the inspection standard registration unit H〇, and the determination reference table 1 1 1 and the like. In the above-described inspection standard setting system 100, the CAD data storage unit 104, the inspection reference database 101, the component shape database 102, the inspection data file 103, and the determination reference table 1 1 1 are each set to the controller 1 Inside the memory 15. The other part is a function set in the control unit 10 by a program installed in the memory 15. This inspection standard sets the CAD data read by the system 100 for the processing of each component and its component type loaded on a particular substrate, or for the processing of the position and size of the terminal block corresponding to each component. Specifically, about each element. 'Settings are used to indicate the assembly position, component type name (square chip, 1C, general name of capacitor, etc.), model, etc., and set information to indicate the position and size of the contact pad (hereinafter referred to as " Terminal block information"). Further, information indicating the position and size of the opening 瓿 of the metal mask used in the solder paste printing step (hereinafter referred to as "mask information") is also set. This mask information was originally attached to the emulsion solder paste printing step. The user of the step is used by the solder height calculating unit 107 in the present embodiment. The above CAD data is read from a CD-ROM set on the CD-ROM drive 16, or transmitted from an external device (not shown), and then taken in by the communication interface 17, and stored in the CAD data memory. Department 1〇4. The inspection reference database 101 and the component shape database 102 are installed in the inspection device in advance by -17- 1358244. However, any of the databases 101, 102 can be appropriately updated by means of a CD-ROM or communication with the outside. In the inspection reference database 101, the details thereof will be described later, and the inspection reference data corresponding to each component type registration and the plurality of material angle shapes which can be generated in the component type will be described. The component types in the database are classified in such a manner as to match the component types of the CAD data, and the same component type names used in the CAD data are assigned to each component type. In the inspection reference data corresponding to a shape of the corner of the material, the data for specifying the color detected from the image of the angle of the drawing is used to specify the data of the program to be measured or determined (for example, the program) The file name or the address indicating the storage location, etc., the rule for determining the threshold for the determination (hereinafter referred to as the "principal limit derivation rule"), etc. The component shape database 102 stores the model number for each component, The component type name, manufacturer name, type, etc. are used to specify the information of the component, and the shape information of the component. The shape data indicates the size or shape of the component by a specific number, and the width of the component body or the height of the component is stored for any component. In addition, there are cases where the unique information is stored for the component type. For example, if it is a component of "1C", the pitch between the wires, the width of each wire, the distance between the opposite ends of the wires, the height of the wires, and the like are stored. The component and the terminal block identifying unit 1〇5 are based on the component type name or model number (hereinafter referred to as "component information") in the CAD data, and the components assembled on the substrate are divided and identified one by one. In addition, for each component, the terminal block corresponding to its component is specified, and the terminal information of each terminal block is obtained. -18-1358244 The window setting unit 106 takes in the terminal information of the corresponding component from the component and the terminal identification unit 1〇5, and sets an inspection target area including the wiring size based on this information (hereinafter referred to as " Wiring block window"). The component information and the terminal block are supplied from the component and the terminal block identifying unit 105 to the inspection reference data search unit 108. The inspection reference data search unit identifies the component type or wiring of the component to be taught from the information, and selects the reference data of the / registered from the inspection reference database 101 for each terminal block. The shape of the skull and read its data. The judgment reference table 111 is used where the inspection reference data is specified. Further, the material height calculation unit 107 performs a process of calculating the height on the terminal block, in addition to a part of the components. When the inspection reference data search unit 108 checks the reference data specified by the inspection reference database 101, the threshold data extraction method in the data is applied to the threshold calculation unit 109, the threshold calculation unit 109, by applying the component data. This threshold is used to calculate the processing rule, and the threshold for the determination of the angle of the material specified for the target component is calculated. The inspection standard registration unit 110 takes in various types of inspection reference data including the determination, and receives information necessary for setting the connection window from the window setting unit 106 (showing the position and size of the window). The information on the "Setting Information of the Terminal Block Window J" is assigned to the information and stored in the inspection data file 1〇3. In addition, the inspection data file is used as the common reference data for each component, and the detection red is also stored. The greening and blue colors required for the lingering limit. The second 値 临 临 値 値 値 値 値 値 . . . . . . . . . . . . . . . . . . . . . . . For the purpose of the welding, the shape is taught to limit the position of the wire holder, and the inspection library -19- 1358244 10 is read and stored in the inspection data file 10 3 by a processing unit (not shown). Further, in the above example, although the CAD data includes information indicating the type of the component, when the CAD data not including the component type information is used, the user inputs the component type information and supplies the input information. Element, terminal block recognition unit 105. In the following, the "square wafer" and "1C", which are the main component types, will be described as an example of the specific content or setting of the inspection reference data. Fig. 3 shows an example of the structure of the data registered in the inspection reference database 1〇1 for the two types of components. In this example, for each component type, the shape of the angle of the material generated by the component type is classified into a plurality of types (hereinafter referred to as "feed angle type"), and individual inspections are set for each type of the angle angle. Benchmark data. In this embodiment, for the convenience of the user, although the names including the name of the component type such as "square wafer A" and "IC_A" are used to indicate the angle type of the material, the angle type is not used to classify the components. It is used to classify the shape of the angle of the material. That is, any element belonging to the same component type also has the possibility of forming a rake angle of all the rake angle types corresponding to the component type. Further, as will be described later, even if the components are the same, there are cases in which the angle patterns are different depending on the terminal block. In the inspection of the reference data, although the above-mentioned various materials are included, in the third figure, it is limited to the threshold and the law is derived. Moreover, in the color glare type inspection apparatus of recent years, although the blue area is always detected, since it is necessary to detect the green and red areas, in this example, it is also desirable to 'measure only the blue area. It also indicates the threshold rule for the measurement of the blue area 値-20 - 1358244. The information in the data shown in Figure 3 is the name of the data angle type or the "image" indicates that each of the data is recorded in the inspection reference database 101. It can also be stored in the inspection standard "feeding angle shape". By means of an optical map of the glare mode, the color corresponding to the corner of the dip angle is blue or red (in Figure 3, the domain is represented by the pattern of the dots to omit the coloration). As the measurement limit of the width and length (the following λ) of the threshold value of the extraction method, since the setting is for the component width, the fourth figure is set according to the component. Mourning. In the case of a square wafer: In the case of the orthogonal direction (vertical direction in the figure), the blue region in the width direction of the device is perpendicular to the element width direction at 1 C. 101 required and threshold 値 derivation rules. "The specific content of the angle type is not necessary. For example, the user can appropriately refer to the file other than the database 101. The pattern of the cross-sectional shape of the noon angle. The pattern of the angle of the material is matched with the actual color distribution, and the red area of the blue area is represented by the pattern of the diagonal line. Because the green area is small, in this embodiment, the setting is used to "Angle width" and "twisting angle rule. In addition, except for a part of the type or the width of the wire multiplied by the established coefficient, the difference between the width of the piece and the length of the element of the length of the material" It is the width of the arrangement of the electrodes. The feed angle width is the maximum 値a of the length of the element, and the length of the feed angle is the maximum 値b of the length of the blue area. But the same, the maximum length of the blue region along the width of the conductor - 21 - 1358244 degrees is taken as the dip angle width, and the length of the blue region along the length of the wire will be The maximum 値 is taken as the length of the feed angle. Here, the relationship between the shape angle of each element type and the threshold 値 derivation rule will be described. First, regarding the "square wafer", although four types of twist angles are set, the square angles of the square wafers A and C are substantially the same. In these versions, a steeper angle of the feed is formed starting from the relatively high position of the component (the upper edge in the illustrated example). The square wafer B is at the same angle as the square wafer A or C. However, since the starting point of the angle of the material is at a relatively low position of the element, the inclination of the angle of the material is slower than that of the square wafers A and C. . Thus, in the image of the angle of the feed, a red area appears on the front end side. The square wafers A and C are distinguished according to the presence or absence of secondary reflection. The secondary reflection means that the reflected light from the dip angle of the adjacent elements located in the opposite relationship is irradiated to the dip angle, and the state of reflection of the illumination light occurs. It is known that secondary reflection is liable to occur when the angle of the coating is steep and the distance from the adjacent angle of the coating is short. In this embodiment, the pattern in which secondary reflection does not occur is referred to as square wafer A, and the pattern in which secondary reflection occurs is referred to as square wafer C. Although the image corresponding to the rake angle of the square wafer A becomes substantially blue in general, the red region caused by the secondary reflection occurs inside the blue region at the image corresponding to the rake angle of the square wafer C. Compared to the above three types, the square wafer D corresponds to a short and steep throw angle because the terminal block is extremely small. The shadow of the pattern angle of this type -22- 1358244 is similar to that of the square wafer A, which becomes substantially blue overall. However, the length of the feed angle is much smaller than that of the square wafer A. The law of the thresholds of each type is derived to reflect the difference in the above images. Specifically, in the square wafer A, it will correspond to 0. The 5 times number is set to the threshold of the feed angle width and will be equivalent to 0 of the component width. The 6-fold number is set to the threshold of the length of the feed angle. On the other hand, in the square wafer B, although the threshold 塡 of the rake angle width is set to be the same as that of the square wafer A, the margin of the rake angle length is set to be larger than the square wafer A in consideration of the existence of the red region. Smaller. Further, regarding the square wafer C, in consideration of the occurrence of secondary reflection, the width and length of the dip angle are set to be smaller than the square wafer A. Regarding the square wafer D having a sharp and short throw angle, the threshold 长 is set to be the same as the square wafer A, but the threshold 塡 of the tantalum angle is fixed to 20 μm. Regarding such a throw angle, it is known empirically to become substantially the same shape regardless of the height of the emulsion solder paste before the angle of the feed. Secondly, in "1C", 1 (:_8) of the 3 type corresponds to a steep angle of the feed. In this type, as with the square wafer, an image in which the substantially whole of the angle of the turn becomes blue is produced. Therefore, the angle of the material corresponding to IC_B is slowly tilted starting from the lower position of the electrode. In the image of the pattern angle of this type, the red area becomes conspicuous. Especially, at the front end corresponding to the angle of the feed. In the portion of the rim, most of the width direction becomes red. The angle of the dice corresponding to IC_C is changed to a shape in which the central portion is flat and the inclined portion -23 - 1358244 is small. This deformation of the sag angle forms a palladium on the surface of the wire. Plating - layer, etc., and the wettability of the wire (the state of soldering the wire) is worse~~. . The situation happened. In the image of the pattern angle of this type, a red area appears along the length direction. Similarly, the marginal derivation rule for "1C" is also made to correspond to the difference in images as shown above, and the thresholds for the determination of each type are different. Specifically, regarding IC_B or IC_C, the determination of the rake angle width is smaller than the threshold value IC_A. Further, regarding the threshold of the length of the feed angle, although the fixed value (50 μm, 30 μm) is set in IC_A or IC_B, the threshold is changed depending on the length of the terminal block with respect to IC_C. In the judgment reference table 111, in order to determine which of the above-mentioned types of the object angle of the teaching object is stored, the information of the content as shown in Fig. 5 is stored. The mode diagram of the "Image" column is for reference and may not be stored in the actual table. In the determination reference table 111, the pattern angle set by the inspection reference database 101 is set to correspond to the height of the solder (hereinafter referred to as "solder height") regarding the formation of the angle of the coating. The solder heights of the square wafers A, B, and C are each expressed as a range of the height of the solder to the height of the element. The solder heights of IC_A, B, and C are expressed in a specific range of 値. On the other hand, with respect to the square wafer D which can also ignore the solder height, zero (nun) 储存 is stored at the solder height. In addition, the type 4 data of the square wafer includes the length of the corresponding wiring base (length of the terminal block). Moreover, the data of the square wafers VIII and c include parameters for discriminating the presence or absence of the secondary reflection (the distance between the junction of the -24-24 - 1358244 wire holder and the adjacent terminal of the opposite component) . In the IC, the data in IC_A and IC_C also contain good and bad information indicating the wettability of the wire. The inspection reference database 101 shown in Fig. 3 and the determination reference table 111 shown in Fig. 5 are connected via the name of the angle pattern. By this, the inspection reference data of each type except for the square wafer D is associated with the solder height range for the formation of the rake angles of the respective shapes. Therefore, if the height of the solder formed by the actual terminal block is known, the inspection reference data corresponding to the height can be specified. Further, in the square wafers A and C, and the A and C of 1C, although given the same height range, according to FIG. 5, the square wafer A and the square wafer C can be distinguished according to the distance from the opposite terminal block. , IC _ A and IC_C can be distinguished according to the good or bad wettability. Further, the height of the solder is not assigned to the corresponding square wafer D, and it can be distinguished from the other three types depending on the length of the terminal. Fig. 6 is a view showing the procedure for determining the angle pattern of the square wafer, and Fig. 7 is a view showing the procedure for the case of determining the angle pattern of 1C. Hereinafter, the details of the determination processing will be described in accordance with the flow of each drawing with reference to Fig. 5. Further, in the drawings and the following description, the steps of the respective processes are simply referred to as ST. In the determination process for the square wafer of Fig. 6, first, for the terminal block in the eye, the length of the terminal block is obtained from the terminal block information in the CAD data, and it is checked whether this is 100^m or more (ST101, 1〇2). . According to Fig. 5, the square wafers A, B, and C correspond to the length of the terminal block of ι〇〇μηι or more -25 - 1358244, but the length of the terminal block corresponding to the square wafer D is less than ΙΟΟμιη. Therefore, in the case where the determination in ST102 is "NO", the state proceeds to ST108' and the tantalum type square wafer d is determined. On the other hand, when the length of the terminal block is ΙΟΟμηη or more, the process proceeds from ST102 to ST103, and the function of the solder height calculating unit 1〇7 is used, and the solder height is calculated by the following calculation processing. First, use the terminal block information or component information in the CAD data to obtain the area of the terminal block or the center position of the component, and read the information of the size of the component body or the electrode from the component shape database 102, and use the information. The area SA of the solder printing target region (the portion not overlapping the electrode) on the terminal block is obtained. In addition, the area SB of the opening of the mask is extracted from the CAD data, and the input of the thickness δ of the mask is received from the user, and the following calculation formula (1) is performed using these. Solder height = (SBxS) / SA . . . . . . . . . (1) In the above formula (1)', it is assumed that the emulsion solder paste embedded in the opening of the mask is melted and uniformly diffused to the entire portion of the terminal which does not overlap with the electrode of the terminal block, thereby obtaining the molten solder. height. Among them, the emulsion solder paste is a method in which a granular solder is charged into a flux, and after melting, the volume of the solder is reduced due to evaporation of the flux, so that the shrinkage rate of the solder corresponding to the reduced amount can be multiplied by the formula (1). Calculate the result as the solder height. Next, in ST1 04, the height of the element of the teaching object is read from the element shape database 102, and it is checked whether the height of the solder calculated in ST103 is 1/3 or more of the element height. According to Fig. 5, the height of the solder height element corresponding to the square wafers a and C is more than 1/3, but the solder height corresponding to the square crystal -26 - 1358244 B is less than 1/3 of the height of the element. Therefore, if the determination of ST104 becomes "NO", the process proceeds to ST 109, and the wafer angle type B is determined. In the case where the height of the solder height element is W3 or more, the process proceeds to ST105. At ST105, the terminal block information in the CAD data is used to calculate the distance from the terminal block of the adjacent component in the relative relationship. Also at ST106, the calculated distance is compared with the predetermined 値LM. According to Fig. 5, although the square wafer A in which secondary reflection has not occurred is at a distance of LM or more from the terminal block, the distance is less than LM in the square wafer C in which secondary reflection occurs. Therefore, if the distance LM calculated in ST103 is equal to or greater than LM in ST106, the process proceeds to ST 107, and the chip angle type square wafer A is determined. On the other hand, if the calculated distance is less than LM, it proceeds to S T 1 1 0, and the tangent angle type square wafer c is determined. Next, after the determination process for 1C, the solder height is calculated, and it is checked whether the height is 0. 05mm or more (ST201, 202). According to Figure 5, IC_A and IC_C correspond to 〇. 焊料 5mm or more solder height, but corresponds to 1 (: _8 solder height is less than 〇. 〇51!1111. Therefore, when 3:202 is ":^0", the process proceeds to ST205, and the throw angle type IC_B is determined. If the solder is highly sturdy. 〇 5 m m or more, proceed to S T 2 0 3 and check that the wettability is good or bad. The information required for this check is entered in advance by the user. For example, a screen for inquiring whether or not to apply the electric charge processing to the 1C mounted on the substrate is displayed, and the answer input to the inquiry is used in the determination of ST203. Here, when it is judged that the wettability is good, the process proceeds to ST204, -27-1358244, and the throw angle type system IC_A is determined. On the other hand, when it is judged that the wettability is not good, the process proceeds to ST206, and the feed angle type IC_C is determined. Next, the entire steps of the teaching process will be described in accordance with FIG. First, at the first ST1, the CAD material of the substrate of the teaching object is read. Next, at ST2, input of information on the thickness of the mask used in the solder printing step or the wettability of the 1C wire is accepted. In ST3, the inspection data file 103 is created. Among them, the inspection data file 103 at this stage only sets an empty file with the file name. In the next ST4, the counter n for counting the number of components is set to start 値1, and thereafter, the loop of ST5 to 14 is executed. In ST5, for the component specified by the counter η (hereinafter referred to as "eye-catching component"), CAD data, specific component type name or model number are used. Further, in ST6, the component shape data of the eye-catching component is read from the component shape database 1 〇 2 according to the model specified in ST5. The read component shape information is stored in a working memory (RAM) in the control unit 10. Next, in ST7, the CAD data is used to specify the position or size of the terminal block corresponding to the eye-catching component. Although not shown in Fig. 8, generally in ST 7, since a plurality of terminal blocks are specified, the following ST8 to 12 are executed for each of the specific terminal blocks. In ST8, a terminal block window including the size of its terminal block is set for each of the specific terminal blocks. In ST9, for each of the terminal blocks, the determination of the angle type of the component depending on the type of the component of the eye-catching element is performed (if the square wafer is the process of Fig. 6, the process 1C is the process of Fig. 7). In -28- 1358244 ST10, for each terminal block, the corresponding inspection reference data determined in ST9 is read. In ST11, for each terminal block, the number of element shapes or wire widths is taken, and these rules are applied to the thresholds in the application check reference data, and the threshold is calculated. In ST12, the setting information of the terminal block window is included, and the setting information of the terminal block window is assigned, and the gun case is 1〇3. At this stage, the set information of the specific wiring materials and the terminal block window are combined, and the set of inspections corresponding to the final form of one component is recorded in the inspection data file 103. Then, the counter n is incremented by 1 at ST 13 . The processing of ST5 to 12 is repeated until the following total number of components. In the above-described procedure, since each of the steps ST8 to 11 is executed at each terminal block, even if the same component is used, the applied inspection reference data may be different. . For example, if the secondary reflection occurs at one of the feeding angles, and the secondary reflection occurs in the other, the square wafer C material is set for the former, and the inspection standard for the latter is set according to the above. Check the relative height of the solder of the reference material according to the size of the component. , the difference in size, etc., for the type of component that is assumed to be a variant of the shape of the angle of the material, the seat of the teaching object can be automatically. Set the inspection reference standard for the specific judgment of the angle of the material type readout component formed in the terminal block of ST9 to be registered in the inspection seat, and check the basic data of each combination. At this point, η exceeds the inspection reference data for which the specified feeding angle is not caused by the terminal block i. The method is different for the shape of the element or the terminal block. The wiring of the components is -29 - 1358244 / Check the reference data. Further, even if the same element is used, if the shape of the gusset is different depending on the substrate to be loaded, the inspection reference data depending on the shape can be set. Further, in the above method, although there is no difference in solder height, in the case where the image of the angle of the dip is different due to the secondary reflection, or the shape of the dip angle is different due to the wettability of the electrode, the response may be set. Inspection reference data for the state of each of the feed angles. Therefore, a high-precision cutting angle inspection can be performed using the set inspection reference data. Further, in the above-described embodiment, although the height of the solder is calculated by calculation, in the emulsion solder paste printing step, the emulsion solder paste applied to each terminal block is inspected, and the solder coating for all the terminal blocks may be performed. A substrate having a good state, inputting the solder height measured at the solder print inspection and inputting the input 値 or multiplying the input 値 by a predetermined coefficient 値' is used as the solder height for forming the angle of the mash ( The coefficient can be determined according to the shrinkage rate and the like accompanying the evaporation of the flux described above). Further, in the above-described embodiment, the dimorphization threshold for detecting the color of the corner portion is used as a common inspection reference material and is set 'but depending on the component, there is also a neighboring portion in the corner portion. The shadow of the component 'and the common two can not cope with the situation. In the case where such a situation may occur, it is preferable to specify a rake angle which is highly likely to be shaded from the distance or positional relationship between the elements, and to set other binarization thresholds at the dip angle. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a view showing the structure of a substrate appearance inspection device according to an embodiment of the present invention. -30- 1358244 Fig. 2 is a functional block diagram showing the inspection reference setting system. Fig. 3 is an explanatory diagram showing a configuration example of an inspection reference database. Fig. 4 is a view showing the complication of the element width, the width of the dip angle, and the length of the dip angle. Fig. 5 is an explanatory diagram showing a configuration example of a determination reference table. Fig. 6 is a flow chart showing the flow of the determination process of the angle pattern for the square wafer. Fig. 7 is a flow chart showing the flow of the determination processing of the angle type for 1C. Figure 8 is a flow chart tm showing the flow of the teaching process for a piece of substrate. [Description of component symbols] 1 Controller 2 Camera 3 Illumination unit 10 Control unit 15 Memory 100 Inspection reference database 102 Component shape database 103 Inspection data file 104 C AD Data memory unit 105 Component % Terminal block identification unit 106 Window Setting unit 1 07 Solder height calculation unit 1354242 108 Inspection reference data search unit 110 Inspection standard registration unit 111 Determination reference table
-32 --32 -